Best Of

I agree that determining the rate of temp rise is crucial, and it's possible I didn't monitor the temp long enough to determine that the rise rate was in fact zero. That 80-minute run was the longest I'd ever observed, and although I thought I'd stood there long enough with the IR gun to confirm zero temp rise for at least a few minutes, that may not have been accurate enough.

The boiler didn't run long enough this morning to get past 145 degree supply temp, so I'll have to wait for colder weather to return in order to get a longer run again.

Meanwhile, I did monitor the supply temp this morning during a 60-minute run. Unfortunately I missed the last 15 minutes of it. But I did collect data manually for 25 minutes, taking one temperature reading per minute. Readings were take from the 1-1/4" copper supply outlet at the boiler, with a handheld IR thermometer aimed at black vinyl electrical tape around the copper. See chart below in the pdf attachment.

The data is somewhat noisy, probably due to turbulent flow at the boiler supply exit. But a linear regression "best fit" done in a Google Sheets chart shows an R-squared of 0.997 for a straight line through the data points, which means the temperature rise is extremely linear and almost perfectly described by a straight line through the data points. This was surprising to me, as I expected it to be more of a curve, bending over as the temperature rose.

The slope of the "best fit" line is 0.80 degrees/minute. So if our 1845 BTU/deg F heat capacity is correct, that implies the cast iron and water are absorbing 0.8 x 1845 = 1476 BTU/min, or 88,560 BTU/hr.

Then from a net output of 124,000 BTU/hr, that leaves 35,440 BTU/hr being radiated. And dividing by a radiator plus pipe area of 554 sq ft gives us an implied radiation of 64 BTU/hr/sq ft.

So let's check the cast iron radiator BTU table. During the above run, I took one surface temp reading about halfway through the run, at the midpoint of a radiator that I expected would represent a system average, being neither the farthest nor the closest rad to the boiler. That surface temp was 124 degrees (which was right about the average water temp at that point, with the supply temp being 135 and the return being 115).

And the radiator BTU table says that, at 124 degrees, a cast iron radiator will output about 60 BTU/hr/sq ft. Which is quite close to our implied output calculation of 64 BTU/hr/sq ft above. So the cast iron radiator BTU chart is correct, and my mistake was in starting from a flawed premise of thermal equilibrium.

So the numbers all fit surprisingly well. The thermal mass is absorbing about 70% of the BTU's, and the surface area is emitting about 30% of the BTU's.

And I think you're right that my mistake was in believing the water temp had stopped rising at 160 F supply, because I didn't monitor long enough. The straight line fit to the temperature rise in this morning's data was a surprise to me, because I had expected more of a flattening of the curve as the temp rises, due to the non-linearity of radiative heat loss. But evidently the rate of change is highly linear.

@EdTheHeaterMan will not be surprised to find that I was wrong again, and that the linearity of the temperature rise means the boiler probably would have hit the high limit eventually, though it would have taken several hours before China Syndrome. 😀 That's good. I learn more from being wrong than being right!

Thanks to everyone who took the time to reply!

It seems like a lot of people have a problem with others trying to understand what is going on around them and how things work.

I don't understand why that is. Learning is important as is understanding how things work and why we do what we do.